вход по аккаунту


Interactive comment on “Moving beyond traditional model - hessd

код для вставки
Hydrol. Earth Syst. Sci. Discuss., 9, C91–C93, 2012
В© Author(s) 2012. This work is distributed under
the Creative Commons Attribute 3.0 License.
Hydrology and
Earth System
9, C91–C93, 2012
Interactive comment on “Moving beyond
traditional model calibration or how to better
identify realistic model parameters: sub-period
calibration” by S. Gharari et al.
Anonymous Referee #1
Received and published: 16 February 2012
In this paper, the authors try to identify model parameters by splitting a calibration
period into several parts, and searching for model parameters that work well for each
of them. The idea is interesting, and the paper deserves to be published. I have a
number of remarks.
Full Screen / Esc
Printer-friendly Version
1) The quality of the English needs improvement. I am not going to list the linguistic
mistakes, because there are so many. Having a native English speaker proofread the
paper would be a great help.
2) Page 1888 line 8: please explain the acronym DYNIA.
3) Top of page 1889: there are actually papers that use entire soil moisture profiles
Interactive Discussion
Discussion Paper
and energy balance data (from Bowen ratio or eddy covariance) to estimate model
parameters (mostly published in AGU journals such as JGR or WRR, some of these
papers go back 10 years or so). This seems to me very relevant to discuss here also.
The methods that the authors develop could and should be applied to multivariate
calibration as well. It may be a different type of models that is used in these studies
("physically-based" models), but that does not mean that they will be better calibrated
than rainfall-runoff models (as a matter a fact, they probably are not, which makes this
methodology all the more relevant to them).
9, C91–C93, 2012
4) I think that the term multivariate is better suited than multicriteria in this respect. A
couple of RMSE values could be minimized, which is all the same criterium, but for
multiple variables. This is just a suggestion, I leave it to the authors to change the
terminology or not.
5) Section 3.2: please state explicitly that the model time step is 12 h.
6) This is a major comment about the method implementation: why did the authors
only use 3 years of data? As the site that they are working is extremely well instrumented, it seems to me hard to believe that only 3 years of data are available. A much
longer model validation would be possible then. I don’t think that the model takes very
much CPU time. So an extra model validation using the obtained parameters should
be possible. What I would suggest is to calibrate the model in the traditional way
(calibration-validation period), and compare the performance of the model using these
parameters during the validation period to the performance using the new method. I
would keep the calibration period the way it is now in the paper (thus the authors do
not need to redo their study), but use the obtained parameters for an extra run with a
validation period of a couple of extra years. That way the benefit of the new method
would really become clear. This would make the paper much stronger. Doing an extra
calibration in the traditional way should be easy for the authors, considering that the
new methodology is much more complicated.
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion Paper
Overall, I think that it is a good paper that should be publised. I urge the authors to take
into account my comments, and especially my last comment (run the model for an extra
couple of years and compare the results to the results from a traditional calibration).
This way the benefit of the new method would be crystal clear. If this extra validation is
not done, it is very well possible that the argument that the new method has not been
validated enough will keep on coming back.
9, C91–C93, 2012
Interactive comment on Hydrol. Earth Syst. Sci. Discuss., 9, 1885, 2012.
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion Paper
Без категории
Размер файла
250 Кб
Пожаловаться на содержимое документа